Cold comfort

Something that has been on my mind for the past week is the political durability of the Albanese government's housing and capital gains tax "reforms". I know they've cleared Parliament, but the housing market has quickly started to cool following the budget of broken promises. And political pressure in Australia can build very quickly.

Falling house prices would be dandy if the cooling was because supply was starting to better respond to demand following reforms to make it more profitable and easier to build new houses, indicating a healthy, welfare-improving adjustment. But the indicators seem to suggest housing prices are falling because, like a knife cutting into both demand and supply, policy is working against profitability, crushing the margins that new stock needs to be built.

Yes, the declines so far are tiny in a historical context. But given how widespread housing ownership is in Australia, it's still the kind of slowdown that tends to be cold comfort for voters because it also slows economic activity. House prices might fall, but affordability doesn't necessarily improve because people start to lose their jobs or their wages fail to keep up with inflation and interest rates, meaning mortgage payments remain relatively high compared to incomes. While the RBA will inevitably respond to slowing economic activity with rate cuts, putting something of a floor under house prices, the 'Bullock put' might be a little late to the party this time around given inflation has been persistently above target for the entirety of her nearly three years in the top job.

The combination of falling house prices, rising rents, a weakening economy, and elevated borrowing costs is politically toxic. Barring some other shock, that pressure will build and could well kill the budget's "reforms" before by next election in May 2028, as it did for the Hawke government's version in the 1980s.

Do we need the ACCC?

You know it's dark days for Australian productivity when the competition regulator uses the shiny new powers (and a $30 million funding boost) granted to it by the Albanese government to block a supermarket (Coles) from building a second site in Kalgoorlie, ostensibly on the grounds of... competition, as the ACCC just did.

Its core justification? That the new Coles store might "lead to the exit of an effective independent competitor", even though it found that "some" consumers were likely to benefit from the new supermarket.

It's clear that the ACCC is conflating consumer welfare with protecting existing competitors from rivalry. It's taking simplistic measures of market power, like static market share counts – which are a poor proxy for actual competitive dynamics – and inevitably flawed predictions about future market structures, then weighing those above acknowledged benefits.

The correct decision here was to allow the market to function; letting Coles build its store and then having consumers decide with their wallets, rather than having a regulator pick winners and losers in Kalgoorlie's supermarket sector.

Alas, that's not how this cookie crumbled. Consumers and competition, which comprise two of the three C's in the ACCC's name, are the victims. The third C, commission, has clearly lost its way. So here's an idea for how to save the federal budget some money: trim its mandate, wind back the new funding and powers the Albanese government gave it, and then look for savings in other places in its nearly half-a-billion dollar annual budget where it's actively harming consumers and competition.

The AI shakedown

The Trump administration is doing its best socialist impression by shaking down various industries for ownership shares. It started with Intel when chips were all the rage and now it has seemingly spread to artificial intelligence, with market leader OpenAI itself suggesting that the US federal government should take a 5% stake in it and every other major US AI lab.

Perhaps OpenAI is just front-running the inevitable, suggesting an amenable 5% to avoid Trump starting at a much higher level. But the cynic in me suspects the motive might be more sinister; that OpenAI wants the government to have a stake because it knows its business model is eventually doomed without it. Really, a 5% stake is a small price to pay for essentially building itself a moat; to have an equity partner in every AI firm with direct power over regulatory policy, including tariffs and other non-trade barriers. After all, the more the government is directly invested in specific companies, the more its incentives are aligned and the more likely it is that those companies will be able to drive up costs for other firms and influence trade policy on "national security" grounds.

The economics of such government equity stakes is poor. At its worst it's socialism; at its best, a form of mercantilism, which was debunked by Adam Smith way back in 1776. It's also unnecessary. If AI is a massive productivity engine, then governments and taxpayers will already capture the upside through higher growth rates and all the goodies that brings, e.g. income and capital gains tax receipts.

My hope is this form of North American populist socialism or mercantilism or whatever it is doesn't spread to Australia, because I'm sure the Albanese government would be all too keen to add an AI Future Fund to its growing portfolio of debt-financed "investment" entities, of which there are now a full dozen. Perhaps the fact that Australia doesn't really have any national champions in the AI space is a blessing; as with everything from cars to solar panels, Australians will be able to buy the cheapest (likely Chinese) models off-the-shelf and not have to worry about governments picking winners and losers for them.

Have a great rest of your weekend.